Performance Analysis of YOLO and Detectron2 Models for Detecting Corn and Soybean Pests Employing Customized Dataset DOI Creative Commons

Guilherme Pires Silva de Almeida,

Leonardo Nazário Silva dos Santos,

Leandro Rodrigues da Silva Souza

et al.

Agronomy, Journal Year: 2024, Volume and Issue: 14(10), P. 2194 - 2194

Published: Sept. 24, 2024

One of the most challenging aspects agricultural pest control is accurate detection insects in crops. Inadequate measures for insect pests can seriously impact production corn and soybean plantations. In recent years, artificial intelligence (AI) algorithms have been extensively used detecting field. this line research, paper introduces a method to detect four key species that are predominant Brazilian agriculture. Our model relies on computer vision techniques, including You Only Look Once (YOLO) Detectron2, adapts them lightweight formats—TensorFlow Lite (TFLite) Open Neural Network Exchange (ONNX)—for resource-constrained devices. leverages two datasets: comprehensive one smaller sample comparison purposes. With setup, authors aimed at using these datasets evaluate performance models subsequently convert best-performing into TFLite ONNX formats, facilitating their deployment edge The results promising. Even worst-case scenario, where with reduced dataset was compared YOLOv9-gelan full dataset, precision reached 87.3%, accuracy achieved 95.0%.

Language: Английский

Intermediates of Hydrogen Peroxide-Assisted Photooxidation of Salicylic Acid: Their Degradation Rates and Ecotoxicological Assessment DOI Open Access
Alicja Gackowska, Waldemar Studziński, Alexander Shyichuk

et al.

International Journal of Molecular Sciences, Journal Year: 2025, Volume and Issue: 26(2), P. 697 - 697

Published: Jan. 15, 2025

Accelerated photooxidation of salicylic acid (SA) was performed using UV radiation and hydrogen peroxide. HPLC-MS analysis showed that the primary intermediates are 2,5-dihydroxybenzoic acid, 2,3-dihydroxybenzoic pyrocatechol, phenol. Deeper oxidation leads to low molecular weight aliphatic acids, such as maleic, fumaric, glyoxylic. The main carried out in same conditions. degradation SA its follows first-order reaction kinetics. In case irradiation alone, photodegradation is slightly faster (reaction rate constant 0.007 min−1) compared (0.0052 min−1). Other products degrade more slowly than SA. Hydrogen peroxide, concentrations 1.8–8.8 mM, accelerates intermediate products. An ecotoxicological evaluation EPI SuiteTM software. overall persistence (POV) long-range transport potential (LRTP) all transformation were assessed OECD POV LRTP screening tool. Salicylic have toxicity. Due their high solubility, these contaminants can travel considerable distances aquatic environment. phenol values 156–190 km. shorter (less 100 km).

Language: Английский

Citations

1

Effects of Acetylsalicylic Acid and Biosolids on Edaphic, Vegetative and Biochemical Parameters of Amelichloa caudata Under Water Shortage Conditions DOI Creative Commons

Julio Molina,

Fernando Silva-Romano,

Irina M. Morar

et al.

Agronomy, Journal Year: 2025, Volume and Issue: 15(4), P. 785 - 785

Published: March 23, 2025

Water scarcity has affected much of Chile for the past 15 years, and Amelichloa caudata, a native species adapted to arid conditions, may offer solution. The hypothesis this study is that both acetylsalicylic acid (ASA) biosolids (BSs) can positively influence plant growth under water stress. This assessed effects ASA BSs on edaphic, physiological, biochemical, productive parameters A. caudata conditions. Results showed treatments enhanced biomass production, height, leaf number, canopy weight. improved retention, mitigating stress leading levels comparable controls. In contrast, did not show significant benefits had lowest values all highest root dry weight was observed in water-restricted plants, while ASA-treated plants lower malondialdehyde (MDA) levels, indicating reduced oxidative However, BS treatment increased MDA suggesting more severe damage. Despite improvements high salt concentrations limit their effectiveness further research required optimize application rates.

Language: Английский

Citations

0

Performance Analysis of YOLO and Detectron2 Models for Detecting Corn and Soybean Pests Employing Customized Dataset DOI Creative Commons

Guilherme Pires Silva de Almeida,

Leonardo Nazário Silva dos Santos,

Leandro Rodrigues da Silva Souza

et al.

Agronomy, Journal Year: 2024, Volume and Issue: 14(10), P. 2194 - 2194

Published: Sept. 24, 2024

One of the most challenging aspects agricultural pest control is accurate detection insects in crops. Inadequate measures for insect pests can seriously impact production corn and soybean plantations. In recent years, artificial intelligence (AI) algorithms have been extensively used detecting field. this line research, paper introduces a method to detect four key species that are predominant Brazilian agriculture. Our model relies on computer vision techniques, including You Only Look Once (YOLO) Detectron2, adapts them lightweight formats—TensorFlow Lite (TFLite) Open Neural Network Exchange (ONNX)—for resource-constrained devices. leverages two datasets: comprehensive one smaller sample comparison purposes. With setup, authors aimed at using these datasets evaluate performance models subsequently convert best-performing into TFLite ONNX formats, facilitating their deployment edge The results promising. Even worst-case scenario, where with reduced dataset was compared YOLOv9-gelan full dataset, precision reached 87.3%, accuracy achieved 95.0%.

Language: Английский

Citations

2